CN107194151B - Method for determining emotion threshold value and artificial intelligence equipment - Google Patents
Method for determining emotion threshold value and artificial intelligence equipment Download PDFInfo
- Publication number
- CN107194151B CN107194151B CN201710262505.9A CN201710262505A CN107194151B CN 107194151 B CN107194151 B CN 107194151B CN 201710262505 A CN201710262505 A CN 201710262505A CN 107194151 B CN107194151 B CN 107194151B
- Authority
- CN
- China
- Prior art keywords
- value
- artificial intelligence
- emotional state
- emotion
- action
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000008451 emotion Effects 0.000 title claims abstract description 131
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 118
- 238000000034 method Methods 0.000 title claims abstract description 39
- 230000002996 emotional effect Effects 0.000 claims abstract description 127
- 230000009471 action Effects 0.000 claims abstract description 97
- 238000012544 monitoring process Methods 0.000 claims abstract description 40
- 238000012545 processing Methods 0.000 claims description 26
- 238000004891 communication Methods 0.000 abstract description 16
- 230000000694 effects Effects 0.000 abstract description 7
- 230000000875 corresponding effect Effects 0.000 description 40
- 230000008859 change Effects 0.000 description 8
- 238000012806 monitoring device Methods 0.000 description 7
- 230000008921 facial expression Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000004590 computer program Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000036760 body temperature Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000003340 mental effect Effects 0.000 description 1
- 230000036651 mood Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000002269 spontaneous effect Effects 0.000 description 1
- 230000000638 stimulation Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/70—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to mental therapies, e.g. psychological therapy or autogenous training
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
- G16Z99/00—Subject matter not provided for in other main groups of this subclass
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Primary Health Care (AREA)
- Epidemiology (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Databases & Information Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pathology (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Multimedia (AREA)
- Human Computer Interaction (AREA)
- Child & Adolescent Psychology (AREA)
- Developmental Disabilities (AREA)
- Hospice & Palliative Care (AREA)
- Psychology (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Computational Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Operations Research (AREA)
Abstract
The application provides a method for determining an emotion threshold value and artificial intelligence equipment, wherein the method comprises the following steps: step 1, determining the emotional state of a first user as a first emotional state by artificial intelligence equipment according to acquired monitoring information; step 2, the artificial intelligence device acquires N actions of a second user; step 3, the artificial intelligence device determines a first action according to the Q value table; step 4, the artificial intelligence device updates the Q value corresponding to the first emotion state and the first action in the Q value table; and 5, the artificial intelligence device determines whether the updated Q value is larger than a preset threshold value, if so, the emotion threshold value is determined according to the monitoring information, and if not, the steps 1 to 5 are repeatedly executed until the emotion threshold value is determined. According to the technical scheme, the emotion threshold value can be optimized by using a Q learning method, so that the communication efficiency and the communication effect under different scenes can be improved.
Description
Technical Field
The present application relates to the field of artificial intelligence and, more particularly, to a method and artificial intelligence device for determining an emotion threshold.
Background
With the continuous development of artificial intelligence, the first task of the new generation of artificial intelligence system is to have emotional connection capability of 'sensibility', so that the common psychological and emotional requirements of people can be met in a way more like real human, and trust and dependence are gradually established. In other words, artificial intelligence should not be just a general mental tool, nor should the path of development of artificial intelligence follow a rational route, but rather should fall within the region of intersection of perceptual and rational.
Emotion calculation plays a crucial role in the development of artificial intelligence, and we see more and more products with "emotion", but this is just a beginning, and artificial intelligence has a long way to go in the face of the extremely complex problem of human emotion.
Currently, the manner of judging the emotion of a user by an artificial intelligence device is to compare an acquired sensor parameter for monitoring physiological information of the user with a preset emotion threshold. For example, if the acquired sensor parameter is greater than the emotion threshold, it may be determined that the emotional state of the user has changed; if the acquired sensor parameter is not greater than the emotion threshold, it may be determined that the emotional state of the user has not changed.
The emotion threshold in the above technical solution is preset in the artificial intelligence device. The emotion thresholds may be different for different users, and the emotion thresholds for the same user in different scenes may also be different. Therefore, how to determine the emotion threshold to be suitable for different users is an urgent problem to be solved.
Disclosure of Invention
The method for determining the emotion threshold and the artificial intelligence device can optimize the emotion threshold, so that the communication efficiency and the communication effect under different scenes can be improved.
In a first aspect, an embodiment of the present application provides a method for determining an emotion threshold, where the method includes: step 1, determining the emotional state of a first user as a first emotional state by artificial intelligence equipment according to acquired monitoring information; step 2, the artificial intelligence device acquires N actions of a second user, wherein the second user is a user communicating with the first user, and N is a positive integer greater than or equal to 1; step 3, the artificial intelligence device determines a first action according to a Q value table, wherein each Q value in the Q value table corresponds to an emotional state and an action, and the Q value corresponding to the first emotional state and the first action is the maximum value of N Q values in the Q value table, wherein the nth Q value in the N Q values corresponds to the first emotional state and the nth action in the N actions, and N is 1, …, N; step 4, the artificial intelligence device updates the Q value corresponding to the first emotion state and the first action in the Q value table; and 5, the artificial intelligence device determines whether the updated Q value is larger than a preset threshold value, if the artificial intelligence device determines that the updated Q value is larger than the preset threshold value, an emotion threshold value is determined according to the monitoring information, and if the artificial intelligence device determines that the updated Q value is not larger than the preset threshold value, the steps 1 to 5 are repeatedly executed until the emotion threshold value is determined, wherein the updated Q value larger than the preset threshold value indicates that the emotional state of the first user is transferred from the first emotional state to a specific emotional state. In the technical scheme, the artificial intelligence device can optimize the emotion threshold by using a Q learning method. Through the optimization of the emotion threshold, the communication efficiency and the communication effect under different scenes can be improved.
With reference to the first aspect, in a first possible implementation manner of the first aspect, the updating, by the artificial intelligence device, the Q value corresponding to the first emotional state and the first action in the Q value table includes: the artificial intelligence device updates the Q value corresponding to the first emotional state and the first action in the Q value table according to the first rate of return.
With reference to the first possible implementation manner of the first aspect, in a second possible implementation manner of the first aspect, the updating, by the artificial intelligence device, the Q value corresponding to the first emotional state and the first action in the Q value table according to the first rate of return includes: the artificial intelligence device updates the Q value corresponding to the first emotional state and the first action in the Q value table using the following formula: qt+1(st+1,at+1)＝(1-λ)Qt(st,at)+λ[rt+γmaxQt(st,at)]Wherein Q ist+1(st+1,at+1) Represents the updated Q value corresponding to the first emotional state and the first action in the Q value table, λ represents the learning strength, Qt(st,at) Representing the Q value corresponding to the first emotional state and the first action in the Q value table before updating, gamma representing a discount factor, rtRepresents the first rate of return, maxQt(st,at) Indicating the maximum Q value corresponding to the first emotional state in the Q value table before updating.
With reference to the first aspect or any one of the foregoing possible implementation manners of the first aspect, in a third possible implementation manner of the first aspect, the method further includes: the artificial intelligence device determines an emotion threshold level; the artificial intelligence device determines the preset threshold according to the emotion threshold level. By setting the emotion threshold level, the optimized emotion threshold can better meet the requirements of users, so that the prediction accuracy can be improved, and the communication efficiency and the communication effect can be improved.
With reference to the first aspect or any one of the foregoing possible implementation manners of the first aspect, in a fourth possible implementation manner of the first aspect, the method further includes: the artificial intelligence device determines an emotion threshold level; the determining the emotion threshold according to the monitoring device parameter corresponding to the first emotion state includes: and determining the emotion threshold according to the emotion threshold level and the monitoring information. By setting the emotion threshold level, the optimized emotion threshold can better meet the requirements of users, so that the prediction accuracy can be improved, and the communication efficiency and the communication effect can be improved.
With reference to the third possible implementation manner of the first aspect or the fourth possible implementation manner of the first aspect, in a fifth possible implementation manner of the first aspect, the determining, by the artificial intelligence device, an emotion threshold level includes: the artificial intelligence device determines the emotion threshold level according to at least one of personalization factor information, session scene information, external environment information, and input information of the first user. The technical scheme takes objective conditions (namely at least one of the personalized factor information, the session scene information, the external environment information and the input information of the first user) into consideration when determining the emotion threshold level, so that the accuracy of prediction can be improved.
With reference to the first aspect or any one of the foregoing possible implementations of the first aspect, in a sixth possible implementation of the first aspect, the method further includes: under the condition that the artificial intelligence equipment determines the emotion threshold value, the artificial intelligence equipment determines whether the current monitoring information is greater than the emotion threshold value; if the current monitoring information is determined to be larger than the emotion threshold, sending indication information, wherein the indication information is used for prompting that if the second user executes the first action, the emotion state of the first user is changed. Therefore, the artificial intelligence equipment can timely remind other users who communicate with the user of avoiding sending actions which can change the emotional state of the user according to the determined emotional threshold value.
In a second aspect, an embodiment of the present application provides an artificial intelligence device, which includes means for performing the first aspect or any possible implementation manner of the first aspect.
In a third aspect, an embodiment of the present application provides an artificial intelligence device, where the artificial intelligence device includes a processor, a memory, and an input device, where the processor is configured to execute instructions stored in the memory, and perform, in conjunction with the memory and the input device, the steps of the first aspect or any one of the foregoing possible implementation manners of the first aspect.
Yet another aspect of the present application provides a computer-readable storage medium having stored therein instructions that, when executed on a computer, cause the computer to perform the method of the above-described aspects.
Yet another aspect of the present application provides a computer program product comprising instructions which, when run on a computer, cause the computer to perform the method of the above-mentioned aspects.
Drawings
FIG. 1 is a schematic flow chart of a method for determining an emotion threshold provided according to an embodiment of the present application.
Fig. 2 is a block diagram of an artificial intelligence device provided according to an embodiment of the present application.
Fig. 3 is a block diagram of an artificial intelligence device provided according to an embodiment of the present application.
Detailed Description
The technical solution in the present application will be described below with reference to the accompanying drawings.
The artificial intelligence device referred to in the embodiments of the present application is any device capable of implementing the method shown in fig. 1, and may be, for example, a computer, a robot, or the like.
FIG. 1 is a schematic flow chart of a method for determining an emotion threshold provided according to an embodiment of the present application.
101, the artificial intelligence device determines initialization parameters, wherein the initialization parameters comprise a Q value table and a preset threshold value.
Each Q value in the Q value table corresponds to an emotional state and an action. For example, table 1 values an illustration of a Q value table.
| A1 | A2 | A3 | A4 | A5 | |
| S1 | Q11 | Q12 | Q13 | Q14 | Q15 |
| S2 | Q21 | Q22 | Q23 | Q24 | Q25 |
| S3 | Q31 | Q32 | Q33 | Q34 | Q35 |
| S4 | Q41 | Q42 | Q43 | Q44 | Q45 |
TABLE 1
Table 1 is a four row five column Q value table. The four rows represent four emotional states, respectively. The four emotional states are S1, S2, S3 and S4, respectively. The five columns represent five actions, respectively. The five actions are a1, a2, A3, a4, and a5, respectively. Each Q value in the Q value table corresponds to an action and an emotional state, e.g., Q11 represents the Q value when the emotional state is S1 and the action is a 1; q12 represents the Q value when the emotional state is S1 and the action is A2, and so on.
Optionally, in an embodiment, in the initial state, all Q values in the Q value table are zero.
And 102, the artificial intelligence device determines the emotional state of the first user as a first emotional state according to the acquired monitoring information.
Specifically, the artificial intelligence device may acquire monitoring information. The monitoring information may be obtained by monitoring physiological, motion, etc. data of the first user by a monitoring device. The monitoring information may be one or more information that can be used to determine the emotional state of the user. For example, the monitoring information may include motion data and/or physiological data. The motion data may include one or more of speech, facial expressions, gestures, standing gestures, and the like. The physiological data may include one or more of heart rate, pulse, skin charge, body temperature, and the like. It is to be understood that if the monitoring information is data that is not represented digitally, such as motion data, the data may be represented digitally to facilitate processing of the data.
The monitoring device is a device, apparatus, or sensor capable of acquiring the monitoring information. The monitoring device may be a plurality of devices or may be one device. For example, the monitoring device may include a device for voice recognition, a device for recognizing facial expressions, a heart rate sensor, a temperature sensor, and the like. For another example, the monitoring device may be a device having a voice recognition function, a facial expression recognition function, and a heart rate and body temperature acquisition function. The monitoring device may be integrated with the artificial intelligence device, or may be an independent device, which is not limited in this application.
The artificial intelligence device may determine the emotional state of the first user directly according to the acquired monitoring information, or may determine the emotional state of the first user after further processing the monitoring information, which is not limited in the embodiment of the present application.
103, the artificial intelligence device obtains N actions of a second user, where the second user is a user communicating with the first user, and N is a positive integer greater than or equal to 1.
The action referred to in the embodiments of the present application refers to all actions uttered by the second user that can be perceived by the first user, including but not limited to voice data, limb actions, facial expression actions, and the like. For example, in a scenario where the first user is conversing with the second user, the artificial intelligence device may retrieve the second user voice data. The speech data may include speech data such as specific content, mood, speed of speech, etc. of the second user speaking. The artificial intelligence device may also obtain a limb movement of the second user.
It will be appreciated that the second user may have one action or may have multiple actions during the course of a conversation.
The artificial intelligence device may obtain the at least one action of the second user in a variety of ways. For example, the artificial intelligence device may have a camera built in to acquire the second user's limb movements, facial expressions, and the like. The artificial intelligence device may also have a built-in microphone to acquire speech data of the second user. For another example, the camera and the microphone may also be external devices, and the artificial intelligence device may acquire the motion and voice data acquired by the external devices.
104, the artificial intelligence device determining a first action according to a Q-value table, wherein each Q-value in the Q-value table corresponds to an emotional state and an action, the Q-value corresponding to the first emotional state and the first action is the maximum of N Q-values in the Q-value table, wherein the nth Q-value in the N Q-values corresponds to the first emotional state and the nth action in the N actions, and N is 1, …, N.
Assuming that the first emotional state of the first user determined by the artificial intelligence device is S1, the N actions of the second user determined by the artificial intelligence device include A1, A2, and A5, Q11 is less than Q12, and Q12 is less than Q15, then the first action determined by the artificial intelligence device is A5.
105, the artificial intelligence device updates the Q value corresponding to the first emotional state and the first action in the Q value table.
The updating the Q value corresponding to the first emotional state and the first action in the Q value table includes: according to a first rate of return, updating the Q value corresponding to the first emotional state and the first action in the Q value table, the Q value corresponding to the first emotional state and the first action, and the Q value corresponding to the first emotional state and the first action.
The artificial intelligence device can determine the first rate of return based on the first action and/or the first emotional state.
Optionally, in some embodiments, the artificial intelligence device may determine the first report back rate according to a preset report back rate table. The rate of return table may include a plurality of rates of return. Each rate of return in the rate of return table corresponds to an action and emotional state. The rate of return in the rate of return package may be determined based on empirical values.
Optionally, in other embodiments, the artificial intelligence device may determine the first rate of return according to a preset formula.
Optionally, in some embodiments, the artificial intelligence device can update the Q value corresponding to the first emotional state and the first action in the Q value table according to the following formula:
Qt+1(st+1,at+1)＝(1-λ)Qt(st,at)+λ[rt+γmaxQt(st,at)](formula 1.1)
Wherein Q ist+1(st+1,at+1) Represents the updated Q value corresponding to the first emotional state and the first action in the Q value table, λ represents the learning strength, Qt(st,at) Representing the Q value corresponding to the first emotional state and the first action in the Q value table before updating, gamma representing a discount factor, rtRepresents the first rate of return, maxQt(st,at) Indicating the maximum Q value corresponding to the first emotional state in the Q value table before updating.
Optionally, in other embodiments, the artificial intelligence device can update the Q value corresponding to the first emotional state and the first action in the Q value table according to the following formula:
Qt+1(st+1,at+1)＝γQt(st,at)+λrt(formula 1.2)
Wherein Q ist+1(st+1,at+1) Indicating the Q value, Q corresponding to the first emotional state and the first action in the updated Q value tablet(st,at) Representing the Q value corresponding to the first emotional state and the first action in the Q value table before updating, gamma representing a discount factor, rtRepresenting the first rate of return.
106. The artificial intelligence device determines whether the updated Q value is greater than a preset threshold, if so, step 107 is executed, and if not, step 102 to step 106 are repeatedly executed until the determination result of step 106 is yes.
And 107, determining an emotion threshold according to the monitoring information.
Optionally, in some embodiments, the preset threshold is determined empirically.
Optionally, in other embodiments, the preset threshold is determined according to an emotion threshold level. The artificial intelligence device may first determine an emotion threshold level and then determine the preset threshold according to the emotion threshold level.
Optionally, in some embodiments, the artificial intelligence device may determine the emotion threshold level based on at least one of personalization factor information of the first user, session context information, external environment information, and input information. According to the technical scheme, objective conditions (namely at least one of personalized factor information, session scene information, external environment information and input information of the first user) are considered when the emotion threshold level is determined, so that the accuracy of prediction can be improved.
And the personalized factor information is used for representing the personalized characteristics of the user. Such as the personality of the user. Specifically, the stimulation of the external things by different people and the stay time of different people in each emotional state can have certain difference between the spontaneous transitions of different emotional states. Optionally, the personalization factor of the first user may be obtained by training the personalization information of the first user according to a basic personalization template library. The personalized factor library can be based on personalized factors of different users obtained by training and learning of a mature basic personalized template library researched by the existing academic world, and the personalized factor library can simulate the emotional state change rules of different types of people, so that the prediction accuracy is improved.
For example, the artificial intelligence device can determine an emotion threshold level based on the first user's personalization factor information. For example, if the first user is not sensitive to external stimuli, the emotion threshold level may be set high; if the first user is sensitive to external stimuli, the emotion threshold level can be set low.
When the first user communicates with different people, the emotion threshold of the first user is different. For example, the first user may be very gently pleasurable when communicating with a small child, and the threshold for triggering a change in the first user's emotional state may be high. For another example, when the first user interacts with some specific objects, the emotional state fluctuation is large, and the threshold value for triggering the first user to change the emotional state may be low. In the embodiment of the present invention, information related to the second user, for example, personalized factor information of the second user, identity information of the second user, and the like, is referred to as session context information.
Different external environments may also affect the emotional state of the tested user. For example, emotional changes of a user to be tested are not the same in a terrorist atmosphere and in a home environment. According to the embodiment of the invention, the information which is used for the emotion prediction module to construct the external environment scene and is irrelevant to the second user is called external environment information. The external environment information may include at least one of the following information: weather information, geographical location information of the first user, and the like. For example, if the weather conditions are good, the artificial intelligence device may set the emotion threshold level to be high; if the weather is poor, the artificial intelligence device can set the emotion threshold level to be low.
In other embodiments, the first user or the second user may directly input the emotion threshold level desired to be set.
Alternatively, in some embodiments, the emotion threshold level may be set high if the artificial intelligence device determines that the relevant information (i.e., personalization factor information, session context information, external context information) is new information that has never been encountered before.
If the artificial intelligence device determines that the session context information is new session context information that has never been encountered before, the emotion threshold level may be set to be lower.
The higher the emotion threshold level is, the higher the preset threshold determined according to the emotion threshold level is; correspondingly, the lower the emotion threshold level is, the lower the preset threshold determined according to the emotion threshold level is. The curve of the change of the preset threshold corresponding to the emotion threshold level may be linear or non-linear, which is not limited in this embodiment of the application.
Optionally, in some embodiments, the correspondence between the emotion threshold level and the preset threshold is predetermined. In other embodiments, the predetermined threshold may be determined based on emotional threshold levels and empirical values.
Each Q value in the Q value table may have a corresponding preset threshold, and the preset thresholds of different Q values may be the same or different. It is understood that the preset threshold in step 106 is the preset threshold corresponding to the Q value of the first emotional state and the first action.
Optionally, in a case that the preset threshold is determined according to the emotion threshold level, the artificial intelligence device may directly determine that the monitoring information is the emotion threshold.
Optionally, in some embodiments, the preset threshold may be the empirical value. In this case, the artificial intelligence device may determine the emotion threshold based on the emotion threshold level and the monitoring information.
For example, the artificial intelligence device determines that the updated Q value is greater than the preset threshold value over M cycles, where M is a positive integer greater than or equal to 2. In this case, the artificial intelligence device may determine that the monitoring information at the M-M cycle is the emotion threshold, where M is a positive integer greater than or equal to 1 and less than M. The value of m can be determined according to the emotion threshold level. If the emotion threshold level is higher, the value of m is larger; the lower the emotion threshold level, the lower the value of m.
For another example, the sentiment threshold is determined based on the monitoring information and a sentiment threshold level. For example, the monitoring information may be multiplied by a coefficient. If the emotion threshold level is higher, the coefficient is higher; the lower the emotion threshold level, the lower the coefficient.
Of course, in some embodiments, the preset threshold and the sentiment threshold may both be determined according to the sentiment threshold level. The detailed description of the embodiments above can be omitted here.
In the technical solution shown in fig. 1, the artificial intelligence device may optimize the emotion threshold by using a Q learning method. Through the optimization of the emotion threshold, the communication efficiency and the communication effect under different scenes can be improved.
In addition, by setting the emotion threshold level, the optimized emotion threshold can better meet the requirements of users, so that the prediction accuracy can be improved, and the communication efficiency and the communication effect can be further improved.
Further, in a case where the artificial intelligence device determines the emotion threshold, the method may further include: the artificial intelligence device determines whether the current monitored information exceeds the sentiment threshold. If so, sending out indication information, wherein the indication information is used for prompting that the emotional state of the first user is changed if the second user executes the first action. Therefore, the artificial intelligence device can remind the second user of which actions can cause the change of the emotional state of the first user in time. The second user can avoid executing corresponding actions to trigger the first user to change the emotional state. The manner of sending the indication information by the artificial intelligence device may be various, for example, an audio prompt, a text prompt, an image prompt, and the like, which is not limited in the embodiment of the present application.
Further, if the first smart device may further determine a difference between the updated Q value and the Q value before updating, if the difference is smaller than a preset difference, it may be determined that the first emotional state in the Q value table and the Q value corresponding to the first action still use the Q value before updating, and if the difference is larger than the preset difference, it may be determined that the Q value corresponding to the first emotional state and the Q value corresponding to the first action in the Q value table is not updated.
Optionally, in some embodiments, if the updated Q value exceeds the preset threshold of the Q value, it may indicate that the emotional state of the user changes from the current emotional state to a specific emotional state. Each change from one emotional state to another has a corresponding preset threshold value of Q value and an emotional threshold value. For example, when the updated Q value exceeds the first preset threshold of the Q value, it may indicate that the emotional state of the user is changing from happy to surprised. For another example, when the updated Q value exceeds the second predetermined threshold of the Q value, it may indicate that the emotional state of the user is changing from surprise to anger. It will be appreciated that the approach shown in FIG. 1 is merely illustrative of one approach to emotion threshold determination. According to the method shown in fig. 1, emotion thresholds corresponding to different emotional state switches can be determined.
The specific emotional state may be an emotional state adjacent to the current emotional state, or may be an emotional state not adjacent to the current emotional state. For example, emotional states may be surprised from happy to angry. The first preset threshold is smaller than the second preset threshold. In some embodiments, the predetermined threshold of the Q value may be the first predetermined threshold and the second predetermined threshold. That is, in these embodiments, the smart device may set an emotion threshold for each emotional state of the first user. In other embodiments, the predetermined threshold of the Q value may be directly set to the second predetermined threshold. In these embodiments, the smart device may set an emotion threshold for only emotional states that need attention (e.g., anger).
Fig. 2 is a block diagram of an artificial intelligence device provided according to an embodiment of the present application. As shown in FIG. 2, the artificial intelligence device 200 can include: a processing unit 201, a storage unit 202 and an acquisition unit 203.
An obtaining unit 203, configured to obtain N actions of the second user, where N is a positive integer greater than or equal to 1.
The storage unit 202 is used for storing a Q-value table.
A processing unit 201 for performing the following steps:
step 1, determining the emotional state of a first user as a first emotional state according to monitoring information;
step 2, acquiring the N actions acquired by the acquiring unit 203, wherein the second user is a user communicating with the first user;
step 3, determining a first action according to the Q value table stored in the storage unit 202, where each Q value in the Q value table corresponds to an emotional state and an action, and the Q value corresponding to the first emotional state and the first action is the maximum value of N Q values in the Q value table, where the nth Q value in the N Q values corresponds to the first emotional state and the nth action in the N actions, and N is 1, …, N;
step 4, updating the Q value corresponding to the first emotional state and the first action in the Q value table stored in the storage unit 202;
and 5, determining whether the updated Q value is larger than a preset threshold value, if so, determining an emotional threshold value according to the monitoring information, and if not, repeating the steps 1 to 5 until the emotional threshold value is determined, wherein the updated Q value larger than the preset threshold value indicates that the emotional state of the first user is transferred from the first emotional state to a specific emotional state.
Optionally, in some embodiments, the processing unit 201 is specifically configured to update the Q value corresponding to the first emotional state and the first action in the Q value table stored in the storage unit 202 according to the first report back rate.
Optionally, in some embodiments, the processing unit 201 is specifically configured to update the Q value corresponding to the first emotional state and the first action in the Q value table stored in the storage unit 202 by using the following formula:
Qt+1(st+1,at+1)＝(1-λ)Qt(st,at)+λ[rt+γmaxQt(st,at)]
wherein Q ist+1(st+1,at+1) Represents the updated Q value corresponding to the first emotional state and the first action in the Q value table, λ represents the learning strength, Qt(st,at) Representing the Q value corresponding to the first emotional state and the first action in the Q value table before updating, gamma representing a discount factor, rtRepresents the first rate of return, maxQt(st,at) Indicating the maximum Q value corresponding to the first emotional state in the Q value table before updating.
Optionally, in some embodiments, the processing unit 201 is further configured to determine an emotion threshold level; the processing unit 201 is further configured to determine the preset threshold according to the emotion threshold level.
Optionally, in some embodiments, the processing unit 201 is further configured to determine an emotion threshold level; the processing unit 201 is specifically configured to determine the emotion threshold according to the emotion threshold level and the monitoring information.
Optionally, the processing unit 201 is specifically configured to determine the emotion threshold level according to at least one of personalization factor information, session context information, external environment information, and input information of the first user.
Optionally, in some embodiments, the artificial intelligence device 200 may further include an output unit, the processing unit 201, and the processing unit is further configured to determine whether the current monitoring information is greater than the emotion threshold in the case of determining the emotion threshold; and if the current monitoring information is determined to be larger than the emotion threshold, indicating the output unit to send out indication information, wherein the indication information is used for prompting that the emotion state of the first user is changed if the second user executes the first action.
The storage unit 202 may also be configured to store information such as emotion threshold level, preset threshold, emotion threshold and the like determined by the processing unit 201
The processing unit 201 may be implemented by a processor. The storage unit 202 may be implemented by a memory. The acquisition unit 202 may be implemented by an input device, such as a microphone, a camera, etc. The output unit may be implemented by an output device, such as a speaker, a display, and the like.
The artificial intelligence device 200 shown in fig. 2 can implement the processes implemented by the method embodiment in fig. 1, and is not described here again to avoid repetition.
Fig. 3 is a block diagram of an artificial intelligence device provided according to an embodiment of the present application. As shown in FIG. 3, the artificial intelligence device 300 can include a processor 301, a memory 302, and an input means 303. The memory 302 may be configured to store information such as a Q-value table, an emotion threshold level, a preset threshold, and an emotion threshold, and may also be configured to store codes, instructions, and the like executed by the processor 301. The various components in the robot 300 are connected by a bus system that includes a power bus, a control bus, and a status signal bus in addition to a data bus.
The artificial intelligence device 300 shown in fig. 3 is capable of implementing the processes implemented by the method embodiment in fig. 1, and is not described here again to avoid repetition.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.
The above-described embodiments may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored on a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, from one website, computer, server, or data center to another website, computer, server, or data center via wire (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (14)
1. A method of determining an emotion threshold, the method comprising:
step 1, determining the emotional state of a first user as a first emotional state by artificial intelligence equipment according to acquired monitoring information;
step 2, the artificial intelligence equipment acquires N actions of a second user, wherein the second user is a user communicating with the first user, and N is a positive integer greater than or equal to 1;
step 3, the artificial intelligence device determines a first action according to a Q value table, wherein each Q value in the Q value table corresponds to an emotional state and an action, and the Q value corresponding to the first emotional state and the first action is the maximum value of N Q values in the Q value table, wherein the nth Q value in the N Q values corresponds to the first emotional state and the nth action in the N actions, and N is 1, …, N;
step 4, the artificial intelligence device updates the Q value corresponding to the first emotion state and the first action in the Q value table;
and 5, the artificial intelligence device determines whether the updated Q value is larger than a preset threshold value, if the artificial intelligence device determines that the updated Q value is larger than the preset threshold value, an emotion threshold value is determined according to the monitoring information, and if the artificial intelligence device determines that the updated Q value is not larger than the preset threshold value, the steps 1 to 5 are repeatedly executed until the emotion threshold value is determined, wherein the updated Q value larger than the preset threshold value indicates that the emotional state of the first user is transferred from the first emotional state to a specific emotional state.
2. The method of claim 1, wherein the artificial intelligence device updating the Q value corresponding to the first emotional state and the first action in the Q value table comprises:
and the artificial intelligence equipment updates the Q value corresponding to the first emotion state and the first action in the Q value table according to a first report back rate.
3. The method of claim 2, wherein the artificial intelligence device updating the Q value corresponding to the first emotional state and the first action in the Q value table according to a first reporting rate comprises: the artificial intelligence device updates the Q value corresponding to the first emotional state and the first action in the Q value table using the following formula:
Qt+1(st+1,at+1)＝(1-λ)Qt(st,at)+λ[rt+γmax Qt(st,at)]
wherein Q ist+1(st+1,at+1) Represents the Q value corresponding to the first emotional state and the first action in the updated Q value table, λ represents the learning intensity, and Qt(st,at) Represents the Q value corresponding to the first emotional state and the first action in the Q value table before updating, gamma represents a discount factor, rtRepresenting said first rate of return, max Qt(st,at) Representing a maximum Q value corresponding to the first emotional state in the Q value table before updating.
4. The method of any of claims 1 to 3, further comprising:
the artificial intelligence equipment determines an emotion threshold level;
and the artificial intelligence equipment determines the preset threshold according to the emotion threshold level.
5. The method of any of claims 1 to 3, further comprising:
the artificial intelligence equipment determines an emotion threshold level;
the determining the emotion threshold according to the monitoring information comprises:
and determining the emotion threshold according to the emotion threshold level and the monitoring information.
6. The method of claim 4, wherein the artificial intelligence device determines an emotion threshold level, comprising:
and the artificial intelligence equipment determines the emotion threshold level according to at least one of personalized factor information, session scene information, external environment information and input information of the first user.
7. The method of claim 1, wherein the method further comprises: under the condition that the artificial intelligence device determines the emotion threshold value, the artificial intelligence device determines whether current monitoring information is larger than the emotion threshold value;
if the current monitoring information is determined to be larger than the emotion threshold value, sending indication information, wherein the indication information is used for prompting that if the second user executes the first action, the emotion state of the first user is changed.
8. An artificial intelligence device, wherein the artificial intelligence device comprises: the device comprises a processing unit, a storage unit and an acquisition unit;
the acquiring unit is used for acquiring N actions of a second user, wherein N is a positive integer greater than or equal to 1;
the storage unit is used for storing a Q value table;
the processing unit is used for executing the following steps:
step 1, determining the emotional state of a first user as a first emotional state according to monitoring information;
step 2, acquiring the N actions acquired by the acquisition unit, wherein the second user is a user communicating with the first user;
step 3, determining a first action according to a Q value table stored in the storage unit, wherein each Q value in the Q value table corresponds to an emotional state and an action, and the Q value corresponding to the first emotional state and the first action is the maximum value of N Q values in the Q value table, wherein the nth Q value in the N Q values corresponds to the first emotional state and the nth action in the N actions, and N is 1, …, N;
step 4, updating the Q value corresponding to the first emotional state and the first action in the Q value table stored in the storage unit;
and 5, determining whether the updated Q value is larger than a preset threshold value, if so, determining an emotion threshold value according to the monitoring information, and if not, repeating the steps 1 to 5 until the emotion threshold value is determined, wherein the updated Q value is larger than the preset threshold value and indicates that the emotional state of the first user is transferred from the first emotional state to a specific emotional state.
9. The artificial intelligence device of claim 8, wherein the processing unit is specifically configured to update the Q value corresponding to the first emotional state and the first action in the Q value table stored in the storage unit according to a first report back rate.
10. The artificial intelligence device of claim 9, wherein the processing unit is specifically configured to update the Q value corresponding to the first emotional state and the first action in the Q value table stored by the storage unit using the following formula:
Qt+1(st+1,at+1)＝(1-λ)Qt(st,at)+λ[rt+γmax Qt(st,at)]
wherein Q ist+1(st+1,at+1) Represents the Q value corresponding to the first emotional state and the first action in the updated Q value table, λ represents the learning intensity, and Qt(st,at) Represents the Q value corresponding to the first emotional state and the first action in the Q value table before updating, gamma represents a discount factor, rtRepresenting said first rate of return, max Qt(st,at) Representing a maximum Q value corresponding to the first emotional state in the Q value table before updating.
11. The artificial intelligence device of any one of claims 8 to 10 wherein the processing unit is further configured to determine an emotion threshold level;
the processing unit is further configured to determine the preset threshold according to the emotion threshold level.
12. The artificial intelligence device of any one of claims 8 to 10 wherein the processing unit is further configured to determine an emotion threshold level;
the processing unit is specifically configured to determine the emotion threshold according to the emotion threshold level and the monitoring information.
13. The artificial intelligence device of claim 11, wherein the processing unit is specifically configured to determine the emotion threshold level based on at least one of personalization factor information, session context information, external environment information, and input information of the first user.
14. The artificial intelligence device of claim 8 wherein the artificial intelligence device further comprises an output unit,
the processing unit is further used for determining whether the current monitoring information is larger than the emotion threshold value under the condition that the emotion threshold value is determined;
and if the current monitoring information is determined to be larger than the emotion threshold, indicating the output unit to send out indication information, wherein the indication information is used for prompting that the emotion state of the first user is changed if the second user executes the first action.
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710262505.9A CN107194151B (en) | 2017-04-20 | 2017-04-20 | Method for determining emotion threshold value and artificial intelligence equipment |
| PCT/CN2018/083885 WO2018192567A1 (en) | 2017-04-20 | 2018-04-20 | Method for determining emotional threshold and artificial intelligence device |
| EP18788158.6A EP3605369B1 (en) | 2017-04-20 | 2018-04-20 | Method for determining emotional threshold and artificial intelligence device |
| US16/657,576 US20200050897A1 (en) | 2017-04-20 | 2019-10-18 | Method for Determining Sentiment Threshold and Artificial Intelligence Device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710262505.9A CN107194151B (en) | 2017-04-20 | 2017-04-20 | Method for determining emotion threshold value and artificial intelligence equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107194151A CN107194151A (en) | 2017-09-22 |
| CN107194151B true CN107194151B (en) | 2020-04-03 |
Family
ID=59871712
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710262505.9A Active CN107194151B (en) | 2017-04-20 | 2017-04-20 | Method for determining emotion threshold value and artificial intelligence equipment |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US20200050897A1 (en) |
| EP (1) | EP3605369B1 (en) |
| CN (1) | CN107194151B (en) |
| WO (1) | WO2018192567A1 (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107194151B (en) * | 2017-04-20 | 2020-04-03 | 华为技术有限公司 | Method for determining emotion threshold value and artificial intelligence equipment |
| US20190385711A1 (en) | 2018-06-19 | 2019-12-19 | Ellipsis Health, Inc. | Systems and methods for mental health assessment |
| EP3811245A4 (en) | 2018-06-19 | 2022-03-09 | Ellipsis Health, Inc. | Systems and methods for mental health assessment |
| CN111797857A (en) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | Data processing method, data processing device, storage medium and electronic equipment |
| US11809958B2 (en) * | 2020-06-10 | 2023-11-07 | Capital One Services, Llc | Systems and methods for automatic decision-making with user-configured criteria using multi-channel data inputs |
| CN113524179A (en) * | 2021-07-05 | 2021-10-22 | 上海仙塔智能科技有限公司 | Control method, device, equipment and medium based on emotion accumulation numerical value |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100825719B1 (en) * | 2005-12-09 | 2008-04-29 | 한국전자통신연구원 | Method for generating emotions and emotions generating robot |
| US9047871B2 (en) * | 2012-12-12 | 2015-06-02 | At&T Intellectual Property I, L.P. | Real—time emotion tracking system |
| US20140191939A1 (en) * | 2013-01-09 | 2014-07-10 | Microsoft Corporation | Using nonverbal communication in determining actions |
| US9386110B2 (en) * | 2014-03-13 | 2016-07-05 | Lenovo Enterprise Solutions (Singapore) Pte. Ltd. | Communications responsive to recipient sentiment |
| US20160055420A1 (en) * | 2014-08-20 | 2016-02-25 | Puretech Management, Inc. | Systems and techniques for identifying and exploiting relationships between media consumption and health |
| CN105868694B (en) * | 2016-03-24 | 2019-03-08 | 中国地质大学（武汉） | The bimodal emotion recognition method and system acted based on facial expression and eyeball |
| CN106295568B (en) * | 2016-08-11 | 2019-10-18 | 上海电力学院 | The mankind's nature emotion identification method combined based on expression and behavior bimodal |
| CN106503646B (en) * | 2016-10-19 | 2020-07-10 | 竹间智能科技（上海）有限公司 | Multi-mode emotion recognition system and method |
| CN107194151B (en) * | 2017-04-20 | 2020-04-03 | 华为技术有限公司 | Method for determining emotion threshold value and artificial intelligence equipment |
-
2017
- 2017-04-20 CN CN201710262505.9A patent/CN107194151B/en active Active
-
2018
- 2018-04-20 WO PCT/CN2018/083885 patent/WO2018192567A1/en unknown
- 2018-04-20 EP EP18788158.6A patent/EP3605369B1/en active Active
-
2019
- 2019-10-18 US US16/657,576 patent/US20200050897A1/en active Pending
Also Published As
| Publication number | Publication date |
|---|---|
| EP3605369A1 (en) | 2020-02-05 |
| US20200050897A1 (en) | 2020-02-13 |
| EP3605369A4 (en) | 2020-03-18 |
| CN107194151A (en) | 2017-09-22 |
| WO2018192567A1 (en) | 2018-10-25 |
| EP3605369B1 (en) | 2024-08-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107194151B (en) | Method for determining emotion threshold value and artificial intelligence equipment | |
| CN110832511A (en) | System and method for learning on distributed devices with data dependent availability | |
| US20150072322A1 (en) | Situated simulation for training, education, and therapy | |
| CN112579909A (en) | Object recommendation method and device, computer equipment and medium | |
| CN108170739A (en) | Problem matching process, terminal and computer readable storage medium | |
| WO2017091629A1 (en) | Reinforcement learning using confidence scores | |
| CN107911491A (en) | Information recommendation method, device and storage medium, server and mobile terminal | |
| CN113240778B (en) | Method, device, electronic equipment and storage medium for generating virtual image | |
| EP2997538A1 (en) | Language proficiency detection in social applications | |
| CN105141587A (en) | Virtual doll interaction method and device | |
| US20200090534A1 (en) | Education reward system and method | |
| US20190070727A1 (en) | Memory control system, system and computer readable medium | |
| CN108959469A (en) | Read management method, device, computer equipment and storage medium | |
| KR20200089588A (en) | Electronic device and method for controlling the electronic device thereof | |
| CN108229640B (en) | Emotion expression method and device and robot | |
| CN111914994A (en) | Method and device for generating multilayer perceptron, electronic equipment and storage medium | |
| CN112925412A (en) | Control method and device of intelligent mirror and storage medium | |
| CN113160819A (en) | Method, apparatus, device, medium and product for outputting animation | |
| CN112949433A (en) | Method, device and equipment for generating video classification model and storage medium | |
| CN113965313A (en) | Model training method, device, equipment and storage medium based on homomorphic encryption | |
| KR20200010061A (en) | Learning method of electronic apparatus and electronic apparatus | |
| CN114115533A (en) | Intelligent interaction method and device | |
| CN117252963A (en) | Digital person generation method and device | |
| CN106926258B (en) | Robot emotion control method and device | |
| CN116468826A (en) | Training method of expression generation model, and method and device for expression generation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |